RED vs. 35mm shoot-out

[Keith Walters]

Hi Max,

 

Most of my money comes from making people look good...

Stephen

And the rest? :lol:

 

But are you suggesting that the point of the exercise is to produce images that are going to look their best where members of the general public are actually going to see them? Whether this is on a movie screen, TV screen or the supermarket flyer that gets stuffed into your letterbox? Images hopefully viewed with approval by randomly chosen and unbiassed consumers?

 

Please don't go there, it will only end in tears, at least on this forum. :lol:

[Keith Walters]

What ASA did you rate the RED at?

 

Mike Brennan

What exactly do they mean by, for example, "320ASA".

 

To me, the only thing it could mean is basically: "If you set the camera to 320ASA, then if you also set your light meter to 320ASA, it should give the correct values for iris, shutter speed, NDs and so on for the lighting conditions you are encountering.

 

Except that (as people quickly discover) in practice it doesn't. Settings that work perfectly well for film can fail miserably with an electronic camera, although the availability if real-time monitoring does at least give a sporting chance of correcting the situation before all the dynamite goes off.

[Michael Gleissner]

What I would be interested in is the MOTION BLUR - has anybody compared capture of fast moving images on RED vs. film?

 

What are the experiences there?

 

MG

[Guest Stephen Murphy]

Interesting stuff. I shot my own tests with RED during the week and posted some of my thoughts here.

[Brian Drysdale]

Interesting stuff. I shot my own tests with RED during the week and posted some of my thoughts here.

 

Just wondering about your post process. Having read a few people's experiences with the RED, it seems to be quite critical and using a traditional HD workflow might not work so well as one making full use of RAW.

[Paul Bruening]

Kodak hater!

[Stephen Williams]

Kodak hater!

 

Hi Paul,

 

I know I am not the only one here! I had to shoot over 30,000' of Kodak recently because the director was not comfortable shooting with Fuji, I wonder what his reaction would have been if I had pushed for digital acquisition?

 

Stephen

[Guest Stephen Murphy]

Hi Brian,

 

All the footage was posted by Screenscene who have already dealt with several commercials shot on RED so are quite familiar with dealing with the footage.

[DJ Joofa]

I'm wondering how that could be. Does 601, for instance, look significantly better to you? In both cases, the gamut covers far more than the region of skin tones in 1931 (x,y) space. With a bit depth of 10 or more, we shouldn't be running out of codes because of the larger gamut (especially towards green) of 709. My guess is it's more likely to be specific implementations -- cameras, monitors, etc.

 

(1) In our extensive development experience with video systems that is a fact we have observed regarding skin tones and Rec 709.

 

(2) The gamuts are not uniform whether Rec 601 or Rec 709. NTSC published extensive material on the non-uniformity of gamuts as far as color fidelity is concerned, which though not directly applicable here, is instructive. Typical white-balance/RGB co-ordinate transformation are defined as linear matrices, and I would tend to think that linear transformation do not capture the non-uniformity equally well everywhere, and specialized non-linear transformations may be needed.

 

Additional complications are imposed by the very non-linear nature of human visual system. According to some evidence vision is not even tri-chromatic everywhere, for e.g., a small enough patch well centered in the field of vision can be matched by mixing only two and not three "primary" colors. Such facts have implications regarding selection of axis regarding the color space.

 

Ref. Willmer and Wright, "Colour Sensitivity of the Fovea Centralis,", Nature, vol. 156, pp. 119-121, July 1945.

 

On the other hand, extensive visual experiments have indeed shown change of color axis (from one set of RGB axis to another set of RGB axis) to be defined by linear transformations. (Recall any linear transformation in a finite dimension space can be defined by a matrix.) Linear theory will stipulate that any three linearly independent vectors suffice to span a 3-D space, and one can transform from one set of basis vectors to others (typically by matrix multiplication).

 

However, we have frequently seen problems that cannot be fully resolved by linear transformations. Could it be an error in our implementations. It is possible, but for many of such things we have not done anything different from what other people do.

 

Again as I said I have a "feeling", and I do not have full evidence.

[Brian Drysdale]

Hi Brian,

 

All the footage was posted by Screenscene who have already dealt with several commercials shot on RED so are quite familiar with dealing with the footage.

 

Hi Stephen,

 

Were the commercials shot with one of the REDs that various Dublin guys had on order?

Edited May 3, 2008 by Brian Drysdale

[Guest Stephen Murphy]

Hi Stephen,

 

Were the commercials shot with one of the REDs that various Dublin guys had on order?

Brian - yup. James Mather has been shooting commercials with his for a few weeks. I was using the Depot's first RED - they have another due for arrival shortly as does James.

[Brian Drysdale]

Brian - yup. James Mather has been shooting commercials with his for a few weeks. I was using the Depot's first RED - they have another due for arrival shortly as does James.

 

Interesting Stephen. They'd be of interest to the feature productions that have been using HDW750s, especially if they've got some RED lenses to keep the glass costs down.

Edited May 3, 2008 by Brian Drysdale

[Max Jacoby]

Again as I said I have a "feeling", and I do not have full evidence.

It's hard to put into words, but I think there is a very obvious difference between the way digital and film render skintones. It's similar with HMI or fluorescents vs tungsten. Tungsten always looks better on skins, HMI and Fluo somehow seem to go into the skin and 'backlighting' the top layer.

[Paul Bruening]

Hey Stephen W.,

 

I love film whether Kodak or Fuji. They're both better than video. I guess I'll drag out my shirt analogy. A cotton shirt is better than a polyester shirt. I can't quantitatively prove that. I think oil paints look better for artwork than acrylic paints. Can't prove that one either. I like the steak at Paulette's better than the steak at Western Sizzlin'. Proof? Nope.

[Guest Glen Alexander]

From the PV website.

 

The Panavision HD-900F camera system is rated at a minimum 320 ISO tungsten.

 

So why would anyone have expectations that this digital camera would be significantly different?

[Van Royko]

I think for this test to be accurate - the two formats need to be color corrected in the same codec and on the same system. Is the davinci making the colour correction off of the scan? If so- the available bandwidth is much greater than pro res 4:2:2 in Color. I think Red cine would provide more latitude from the footage than pro res and Color.

 

There is however, very little doubt in my mind that more highlights can be recovered from film.

 

 

I agree totally on the lens issue, when LACC wants to pay for a full 4k test, I'll rent matching Master Primes.

 

However, for something put together on an LACC budget (it's the best film school secret in LA, but they have very limited resources), I think the lens's are a close enough in quality for an HD test (or, more likely, an H.264 SD test, since I imagine that is most of what you are looking at). But I deliberately avoided shooting any lens or sharpness charts, since the lens quality difference seemed to be to make those tests moot.

 

The footage is in AppleProRes (HQ) Rec. 709 from REDcine. After talking to some RED owners, it seems like I should expect a little more latitude if I export in REDlog, so I will attempt to re-export in that. All timing down in COLOR or DA VINCI was aiming for a neutral (blacks balanced and whites balanced) grade with blacks around 7 and whites around 100 with a good stretch to the gamma.

 

The workflow chosen was the workflow recommended by RED at RED DAY as being what they would like to call a "common" one; they were careful not to recommend any specific workflow (one of their pluses is that there is so much flexiblity), but they also said that they expected Rec. 709 to be a common workflow.

 

My students have asked me to play with the RED footage to see if I can get it to saturate without noise the way the film does; I will work on it and post.

 

thanks for the thoughts.

 

Ch:H

[Nate Downes]

From the PV website.

 

The Panavision HD-900F camera system is rated at a minimum 320 ISO tungsten.

 

So why would anyone have expectations that this digital camera would be significantly different?

Because that is what many of the RED fans kept proclaiming. I remember one going "It can do 10,000 ISO" over and over again.

[Stephen Williams]

Because that is what many of the RED fans kept proclaiming. I remember one going "It can do 10,000 ISO" over and over again.

 

Hi,

 

Yes indeed there were many claims.

 

I prefer to test for myself to see what works, then I decide if I can use the equipment or not.

 

Stephen

[Thomas James]

I think the problem with 35mm film is that it is too grainy to be considered ultra high definition. I saw the HD-DVD version of the 1966 movie Grand Prix which was originally shot on 65mm film at it looked great because it was sharp clear and clean. Of course Red does not have the resolution of 65mm film but it matches the cleaness of the 65mm film format. With the introduction of the Epic it will be interesting to see if digital will be a real alternative to 65mm film production.

[John Atala]

Thank you guys, and for all the collegues who donated their equipment for the elaboration of this literature! Thanks!

[georg lamshöft]

"I think the problem with 35mm film is that it is too grainy to be considered ultra high definition"

 

My experience from still photography: Use the best of both worlds!

 

I'm shooting slides (usually Velvia 50ASA or 100F) and always thought that a chemical/optical cibachrome print (easily >50$ each) is the best way to present them. I was disappointed, the cibachrome prints were brilliant (the typical cibachrome glossy, strong colors), but they also looked grainy and soft in comparison to digital prints from my Leica M8 (by the way with a quite 4k-35mm-like sensor with 27mm width ~4000 pixels and 6,8µm pixel pitch).

Then I scanned the slides with a 6300dpi Imacon (~ the resolution of a 6k Arriscan?) and used all the digital tools (sharpening, anti-grain/noise tools) and printed them. I was surpised, the digital print seemed as "grainy" and even a bit sharper than the cibachrome made from a 6x6-slide (Zeiss 80mm - not as sharp as the Leica - but still 3,5x image area) and had more detail than the 10MPixel-M8-File! It still looked somehow film-like, no typical digital artifacts and very low grain - from now on I don't use cibachrome anymore and still love shooting 35mm slides for big prints!

 

To my understanding it's the same with film/35mm/DI:

http://www.cinematography.com/forum2004/in...amp;hl=arriscan

 

Whenever we compare systems we should look at the whole workflow, maybe the ideal 35mm-solution is using modern DI-technology (6k->4k?) and profit from all the digital tools like grain-reduction and then compare it to other systems instead of making a war "100% digital" vs. "100% chemical/optical"?

 

I hope you got the idea :rolleyes:

Edited May 4, 2008 by georg lamshöft

[Max Jacoby]

I think the problem with 35mm film is that it is too grainy to be considered ultra high definition.

That's bullshit Thomas and you know it.

[Brian Drysdale]

The 35mm to 70mm blow ups for flagship cinemas were always impressive compared to the normal 35mm release prints. So there's a lot more there than you see in your average multiplex screening.

[Max Jacoby]

Whenever we compare systems we should look at the whole workflow, maybe the ideal 35mm-solution is using modern DI-technology (6k->4k?) and profit from all the digital tools like grain-reduction and then compare it to other systems instead of making a war "100% digital" vs. "100% chemical/optical"?

I really don't understand why people are so averse to grain. I for one love grain, it gives the images a lovely texture that digital lacks. If you want to avoid grain, just shoot anamorphic. Really, modern filmstocks are so low in grain anyway that it is not really an issue even if you shoot spherical.

[John Sprung]

What I would be interested in is the MOTION BLUR - has anybody compared capture of fast moving images on RED vs. film?

Yes -- search this site for the threads on the "Rolling Shutter".

 

 

 

-- J.S.